Draft a minimal Windows VNC viewer from a brief
They asked Claude to build a tiny Windows 10 VNC client from an email spec, but after a week it still did not work.
- What they did
- The person emailed Claude a short specification for a stripped-down x64 VNC viewer on Windows 10. It had to reach localhost:5921 through an SSH tunnel, display only the remote pointer, and use the highest quality setting on a gigabit LAN. Claude was supposed to compile it on a Mac in Rust or C++, then share the binary, a SHA-256 checksum, and a PowerShell command for checking it.
- How it went
- The attempt did not succeed. Claude kept returning unusable builds, exhausted the full 20x Max weekly limit, and was still sending broken code even after the reset three days later. The person shut down the project after roughly a week.
- Tips
- For a client that speaks a specific protocol, expect several repair cycles and decide in advance when to stop. Only ask for the checksum and verification command once the executable actually runs.
Try it yourself with Claude Code
Create a very small [app type] for [target OS and version] that reaches [host and port] only through [tunnel or proxy details]. I’m developing on [my OS], so use cross-compilation if needed and avoid unnecessary dependencies. First outline the build plan and wait for my approval before anything irreversible, such as submitting, deleting, or releasing files. When the build is finished, provide the executable, its SHA-256 checksum, and the exact command to verify it.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Move a Chrome extension into Safari
A developer had ChatGPT rewrite Session Buddy from Chrome for Safari and get it working in roughly 20 minutes.
Use spare AI quota to draft reviewable PRs
A founder pointed an agent at goals and task lists so it could turn leftover Claude and Codex quota into pull requests for review.
Use parallel agents to build a compiler
Coordinate many coding agents in containers and back them with tests and an oracle while they work on one compiler repo.
Have agents co-build a reference system with role splits
A solo founder divided an AI-assisted build into roles and shipped a 120-model mental-models framework with an MCP server.
Build a PR dashboard across multiple repos
They used ChatGPT for planning and an agent in WebStorm to assemble a dashboard that tracked pull requests across four Azure DevOps repos.
Use an AI spec to brief a coding agent
Work out a feature with an AI assistant, then pass the spec to a coding agent and verify the result by testing.